By Hellinger Software.

Class to handle text files as memory mapped files.

Including efficient methodes for access like sequentiell read

or random access read to text, high performance search routines

and many more.

{
==============================================================================
MapTextfiles - Release 99/1 14.03.1999
for Delphi 4, should also run with Delphi 3 ------------------------------------------------------------------------------ MapTextfiles is FREEWARE. Freeware means, that you can use this software
without paying anything for it, but it is not public domain! This software
is protected through the law of the Federal Republic of Germany, the European
Union and other countries. (C)1999 by Peter Hellinger Software, All rights are reserved. Peter Hellinger Software, Zerzabelshofstrasse 41, D-90480 N黵nberg email: mail@helli.de
homepage: http://www.helli.de ============================================================================== Installation:
Copy MAPTEXTFILE.PAS into your library path. Use ist. 8-) ------------------------------------------------------------------------------ In Delphi you have many ways to manipulate text files. But all of these have
a handicap: You must read the whole file (StringList) or you can only access
data sequential (Assign/Readln). The Windows API provides the so called Memory Mapped Files. Windows self
uses MMFs to load EXE or DLL. Therefore the mechanism is very efficient, but
simple to handle. The only handicap is, that you must know the size of the
file before you access it. This means for typical text files, that the
operation is normally limited to read from the file or manipulate inside. The class TMapTextfile wraps the neccesary API calls and povides efficent
functions for accessing the data. TMapTextfiles provides the following properties and methods: Methodes:
========= Create Creates an instace of the class Destroy Destroys the instance Open Opens a file as a memory mapped file. filename = name of the file
mode = open mode:
mmRead = Open only for read
mmReadWrite = Open for read and wrie Returns INVALID_HANDLE_VALUE if the file not exist.
If the result > 0 all is OK. NOTE:
1. The file must exist!
2. You cannot write behind the end of the file. Close Close the memory mapped file and frees all handles. EndOfFile Returns TRUE, if the End of the file is reached. GetSize Returns the Size of the file in Bytes GetPos Returns the actual file read/write position
NOTE: Position 0 is the start of the file SetPos Sets the actual read/write position
NOTE: Position 0 is the start of the file ReadChar Reads a character from the actual read/write position.
The r/w position is incremented. ReadString Returns a string, starting at the actual r/w position.
The string is delimited by characters < SPACE, but not
by ESC (#27) and TAB. The r/w position moves to the
end of the string, delimiter chararcters are skiped. ReadLn Same as ReadSring, but as a Procedure. ReadCharAt Reads a charater from an arbitray possition.
The r/w position is NOT moved! pos = position to read from (0 = start of the file!) ReadChars Reads a line of charactes from the MMF.
The r/w position is NOT moved! str = The buffer to read to
pos = position to read from (0 = Start of the file!)
len = number of characters to read. ReadStringAt Returns a string, starting at an arbitray possition.
The string is delimited by characters < SPACE, but not
by ESC (#27) and TAB. The r/w position is NOT moved. FindString Findes a substring in the MMF and Returns the position. str = rhe substring to search for
pos = position to start the search (0 = start of the file)
max = position to end the search. If this is 0 or less
then 0 the end of the file is the limit. Returns the position of the substring or -1 if the
substring is not found. FindWildCard Same as Findstring, but supports wildcard search. str = the substring to search for
pos = position to start the search (0 = start of the file)
max = position to end the search. If this is 0 or less
then 0 the end of the file is the limit.
wild = the character used as wildcard (i.e. "*")
joker = the character used as joker (i.e. "?") Returns the position of the substring or -1 if the
substring is not found. ReadBytes Reads a number of bytes to a anonymous variable.
The r/w position is NOT moved! b = the anonymous variable
pos = position to read from (0 = start of the file)
len = number of bytes to read. WriteBytes Writes a number of bytes to the file.
NOTE: You can not write behind the end of the file!!! b = the anonymous variable
pos = position to write to (0 = start of the file)
len = number of bytes to write ==============================================================================
} unit uMapTextfile; interface uses Classes, Windows; type
tMapMode = ( mmRead, mmReadWrite ); type
TMapTextfile = class
private
f_file : THandle;
f_MMF : THandle;
f_size : INTEGER;
f_view : PByte;
f_data : PChar;
f_pos : INTEGER;
f_open : BOOLEAN;
function CalcPos( pos : INTEGER ) : PChar;
public
constructor Create;
destructor Destroy; override;
function Open( const filename : string; mode : tMapMode ) : INTEGER;
procedure Close;
function ReadChar : CHAR;
function ReadString : string;
procedure ReadLn( var str : string );
function ReadCharAt( pos : LONGINT ) : CHAR;
procedure ReadChars( str : PChar; pos, len : LONGINT );
function ReadStringAt( pos : LONGINT ) : string;
function GetSize : LONGINT;
function GetPos : LONGINT;
procedure SetPos( pos : LONGINT );
function EndOfFile : BOOLEAN;
function FindString( const str : string; pos, max : INTEGER ) : INTEGER;
function FindWildCard( const str : string; pos, max : INTEGER;
wild, joker : CHAR ) : INTEGER;
procedure ReadBytes( var b; pos, len : LONGINT );
procedure WriteBytes( var b; pos, len : LONGINT );
end; implementation constructor TMapTextfile.Create;
begin
f_open := FALSE;
end; function TMapTextfile.Open( const filename : string; mode : tMapMode )
: INTEGER;
var
m1, m2, m3 : CARDINAL;
begin
f_open := FALSE;
if mode = mmRead then
begin
m1 := GENERIC_READ;
m2 := PAGE_READONLY;
m3 := FILE_MAP_READ;
end else begin
m1 := GENERIC_READ + GENERIC_WRITE;
m2 := PAGE_READWRITE;
m3 := FILE_MAP_WRITE;
end; f_file := CreateFile( PChar( filename ), m1, FILE_SHARE_READ, nil,
OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, ); if f_file = INVALID_HANDLE_VALUE then
begin
Result := INVALID_HANDLE_VALUE;
EXIT;
end;
try f_size := GetFileSize( f_file, nil ); f_MMF := CreateFileMapping( f_file, nil, m2, , f_size, nil ); if f_MMF = then
begin
Result := -;
EXIT;
end;
finally CloseHandle( f_file );
end; try
f_view := MapViewOfFile( f_MMF, m3, , , f_size );
if f_view = nil then
begin
Result := -;
EXIT;
end;
finally
CloseHandle( f_MMF );
end;
f_data := PChar( f_view );
f_pos := ;
f_open := TRUE;
Result := ;
end; destructor TMapTextfile.Destroy;
begin
if f_open then
Close;
inherited;
end; procedure TMapTextfile.Close;
begin
if f_open then
begin
UnmapViewOfFile( f_view );
f_open := FALSE;
end;
end; function TMapTextfile.CalcPos( pos : INTEGER ) : PChar;
begin
Result := nil;
if f_open then
begin
if pos < then
pos := ;
if pos > f_size then
pos := f_size;
Result := PChar( LONGINT( f_view ) + pos );
end;
end; function TMapTextfile.ReadChar : CHAR;
begin
Result := #;
if f_open then
begin
f_data := PChar( LONGINT( f_view ) + f_pos );
Result := f_data^;
INC( f_pos );
end;
end; function TMapTextfile.ReadString : string;
begin
Result := '';
if f_open then
begin
f_data := PChar( LONGINT( f_view ) + f_pos );
while ( f_pos < f_size ) do
begin
case f_data^ of
# .. # :
case f_data^ of
#, # :
Result := Result + f_data^; // Tab und Escape weiterreichen
# :
begin
INC( f_pos );
EXIT;
end;
# :
begin // Carriage Return terminiert
INC( f_pos );
INC( f_data );
if f_data^ = # then
INC( f_pos );
EXIT;
end;
end;
else
Result := Result + f_data^;
end;
INC( f_pos );
INC( f_data );
end;
end;
end; function TMapTextfile.ReadCharAt( pos : LONGINT ) : CHAR;
begin
if f_open then
Result := CalcPos( pos )^
else
Result := #;
end; procedure TMapTextfile.ReadChars( str : PChar; pos, len : LONGINT );
var
i : INTEGER;
p : PChar;
begin
if f_open then
begin
if len <= then
EXIT;
i := ;
p := CalcPos( pos );
while ( i <= f_size ) and ( i <= len ) do
begin
str^ := p^;
INC( str );
INC( p );
INC( i );
end;
end;
end; procedure TMapTextfile.ReadBytes( var b; pos, len : LONGINT );
var
p : PChar;
begin
if f_open then
begin
p := CalcPos( pos );
Move( p^, b, len );
end;
end; procedure TMapTextfile.WriteBytes( var b; pos, len : LONGINT );
var
p : PChar;
begin
if f_open then
begin
p := CalcPos( pos );
Move( b, p^, len );
end;
end; function TMapTextfile.ReadStringAt( pos : LONGINT ) : string;
var
i : INTEGER;
p : PChar;
begin
Result := '';
if f_open then
begin
p := CalcPos( pos );
i := ;
while ( i <= f_size ) do
begin
case p^ of
# .. # :
case p^ of
#, # :
Result := Result + p^; // Tabs und Escape weiterreichen
#, # :
EXIT; // Linefeed and Carriage Return terminiert
end;
else
Result := Result + p^;
end;
INC( p );
end;
end;
end; procedure TMapTextfile.ReadLn( var str : string );
begin
str := ReadString;
end; function TMapTextfile.GetSize : LONGINT;
begin
if f_open then
Result := f_size
else
Result := -;
end; function TMapTextfile.GetPos : LONGINT;
begin
if f_open then
Result := f_pos
else
Result := -;
end; procedure TMapTextfile.SetPos( pos : LONGINT );
begin
if f_open then
begin
if pos < then
pos := ;
if pos > f_size then
pos := f_size;
f_pos := pos;
end;
end; function TMapTextfile.EndOfFile : BOOLEAN;
begin
if f_open then
Result := f_pos >= f_size
else
Result := TRUE;
end; function TMapTextfile.FindString( const str : string; pos, max : INTEGER )
: INTEGER;
var
s, l1, j : INTEGER;
p, x : PChar;
begin
Result := -;
if f_open then
begin
if max <= then
max := f_size;
if pos < then
pos := f_pos;
if pos > max then
EXIT;
x := PChar( str );
p := PChar( f_view );
l1 := ;
while ( x[ l1 ] > # ) do
INC( l1 );
if ( l1 > ) then
begin
s := pos;
repeat (* 1 *)
j := ;
while ( s + j < max ) and ( j < l1 ) and ( x[ j ] = p[ s + j ] ) do
begin
INC( j );
if ( j = l1 ) then
begin
Result := s;
EXIT;
end;
end;
INC( s );
until s >= f_size;
end;
end;
end; function TMapTextfile.FindWildCard( const str : string; pos, max : INTEGER;
wild, joker : CHAR ) : INTEGER;
var
s, l1, j : INTEGER;
p, x : PChar;
begin
Result := -;
if f_open then
begin
if max <= then
max := f_size;
if pos < then
pos := f_pos;
if pos > max then
EXIT; x := PChar( str );
p := PChar( f_view );
l1 := ;
while ( x[ l1 ] > # ) do
INC( l1 ); if ( l1 > ) then
begin
s := pos;
repeat (* 1 *)
j := ;
while ( s + j < max ) and ( j < l1 ) and
( ( x[ j ] = p[ s + j ] ) or ( x[ j ] = joker ) ) do
begin
INC( j );
if ( x[ j ] = wild ) or ( j >= l1 ) then
begin
Result := s;
EXIT;
end;
end;
INC( s );
until s >= f_size;
end;
end;
end; end.

TMapTextfile v.99/1的更多相关文章

  1. 关于更新发布CSS和JS文件的缓存问题

    现如今,浏览器大战下,各个浏览器也是拼了命的提高性能,升级JS解析引擎,更好的处理浏览器的页面缓存,让用户的浏览体验更快,占用更小的PC资源.那么,问题就出现在JS和CSS缓存,甚至页面缓存上.至于浏 ...

  2. java基础学习总结-接口

    原文链接:http://www.cnblogs.com/xdp-gacl/p/3651121.html 一.接口的概念 JAVA是只支持单继承的,但现实之中存在多重继承这种现象,如"金丝猴是 ...

  3. aes加密C语言

    /** * \file aes.h * * \brief AES block cipher * * Copyright (C) 2006-2010, Brainspark B.V. * * This ...

  4. CentOS_7 OpenWrt Eclipse 环境搭建与 Dr.com 开发笔记

    一:内核的编译. 1,linux 编译环境的搭建与源码的准备工作 2,常用软件的安装 (make menuconfig) 3,  针对TP-Link WR740N 一些软件的openwrt 的移植 4 ...

  5. 转:小白编译openwrt固件教程

    原文地址 编译openwrt固件并没有想象的那么复杂,我也是个小白,以下内容是我将网络上的编译教程稍微进行了一下整合.因为我发现很多编译教程没有说明如何更改flash相关配置.   安装ubuntu, ...

  6. Openwrt Image Builder/SDK 初探

    image builder和SDK既可以从官网上下载,又可以自己进行编译(make menuconfig).官网上下载的是预先帮你编译好的,这样可以大量节省自己编译源码花的时间,这两个东西相当于半成品 ...

  7. 搭建自己的OpenWrt开发环境

    1.  安装环境Linux系统,如果在CentOS上操作,需安装如下依赖包:yum install binutils bzip2 gawk gcc gcc-c++ gettext makencurse ...

  8. js文件被浏览器缓存的思考

        我们的用户量大,修改js文件后,用户反馈登录出现问题.实际上刷新一下就没事了.就是因为用户的浏览器使用的还是本地缓存的js代码.   强制刷新一般就会重新去服务器获取新的js代码.但不能让用户 ...

  9. Openwrt 初探

    最近想研究一下Openwrt,于是开始搭建openwrt环境,虽然现在没有现成的板子,但是 可以先编译起来. openwrt的特点是基于下载 -> patch -> 编译 的一个工作模式, ...

随机推荐

  1. set -o vi AIX下shell

    set -o vi 再用esc+K键就可以使用上一条指令了 esc+kesc+j上下翻 ksh默认是emacs风格的.set -o emacs 在AIX下使用自己已经使用过的命令 在AIX下使用,默认 ...

  2. 图像显著性论文(一)—A Model of saliency Based Visual Attention for Rapid Scene Analysis

    这篇文章是图像显著性领域最具代表性的文章,是在1998年Itti等人提出来的,到目前为止引用的次数超过了5000,是多么可怕的数字,在它的基础上发展起来的有关图像显著性论文更是数不胜数,论文的提出主要 ...

  3. git - git命令中文显示乱码

    使用git add添加要提交的文件的时候,如果文件名是中文,会显示形如897\232\350\256...的乱码,解决办法:   git config --global core.quotepath ...

  4. 取消cp确认

    使用cp -r -f 强制覆盖拷贝命令时,每一个文件都需要认为的键入“Y”进行确认 可以通过取消同名cp的定义,命令如下unalias cp

  5. hdu4307

    好题,详细题解在这里http://blog.csdn.net/weiguang_123/article/details/8077385 这里回顾一下: 当i和j都在一个集合里会产生新的收益,是经典题直 ...

  6. day4 递归二分法查找

    现有一个序列,data=[for i in range(1,5000,3)],现在要求看一个数是否在列表中存在,我们知道,我们可以使用in或__contains__()的方法,判断一个值是否在列表中, ...

  7. day3修改配置文件

    有如下配置文件,在指定文件位置添加一条新的记录: global log 127.0.0.1 local2 daemon maxconn log 127.0.0.1 local2 info defaul ...

  8. js_数组对象的浅克隆

      如果再考虑更奇葩更复杂的情况,例如我们定义: var obj = [{ "a": { "a1": ["a11", "a12&q ...

  9. 洛谷P3052 [USACO12MAR]摩天大楼里的奶牛 [迭代加深搜索]

    题目传送门 摩天大楼里的奶牛 题目描述 A little known fact about Bessie and friends is that they love stair climbing ra ...

  10. Android之 解析XML文件(1)—— Pull解析

    (以下文章基本照抄<第一行代码>) 解析XML文件有很多方法,这里主要讲Pull解析和SAX解析.这篇文章主要是讲Pull解析. 一.Pull解析参考代码 先上代码: private vo ...